利用生成对抗网络的时频图像去噪和增强处理  

Time-Frequency Image Denoising and Enhancement Processing Based on Generative Adversarial Network

作　　者：李昆 朱卫纲 LI Kun;ZHU Weigang(Department of Graduate Management,Space Engineering University,Beijing 101416,China;Department of Electronic and Optical Engineering,Space Engineering University,Beijing 101416,China)

机构地区：[1]航天工程大学研究生管理大队,北京101416 [2]航天工程大学电子与光学工程系,北京101416

出　　处：《电讯技术》2020年第5期517-523,共7页Telecommunication Engineering

基　　金：电子信息系统复杂电磁环境效应(CEMEE)国家重点实验室项目(2018Z0202B)。

摘　　要：针对雷达信号时频图像的去噪和增强问题,提出了利用生成对抗网络二次生成时频图像的方法。首先利用时频分析产生雷达信号的时频图像作为原始数据集1;接着利用生成对抗网络对数据集1进行学习之后生成新的数据集2,数据集2相对于数据集1拥有着去噪和增强的效果;最后提取时频图像奇异值特征检验生成的数据集2的有效性。对6种常见的雷达信号的时频图像进行了仿真实验,结果证明了该方法在时频图像去噪和增加样本多样性方面是有效的。To deal with the problem of denoising and enhancement of radar signal time-frequency images,a method of secondarily generating time-frequency images by generative adversarial network is proposed.Firstly,time-frequency analysis is used to generate the time-frequency image of the radar signal as the original data set 1.Then,after learning the data set 1 by using the generative adversarial network,a new data set 2 is generated,and the data set 2 has denoising and enhancement effects relative to data set 1.Finally,the validity of the data set 2 generated by the time-frequency image singular value feature is checked.Experiments on the time-frequency images of six common radar signals are carried out.The results show that the method is effective in time-frequency image denoising and increasing sample diversity.

关 键 词：雷达辐射源识别 时频图像去噪 生成对抗网络 奇异值分解 

分 类 号：TN971[电子电信—信号与信息处理]